THEMIS-PCI: Patients with Diabetes and Stable Coronary Artery Disease with a History of Prior Percutaneous Coronary Intervention Randomized to Ticagrelor or Placebo

Deepak L Bhatt,* Philippe Gabriel Steg,* Shamir R Mehta, Lawrence A Leiter, Tabassome Simon, Kim Fox, Claes Held, Marielle Andersson, Anders Himmelmann, Wilhelm Ridderstråle, Jersey Chen, Yang Song, Rafael Diaz, Shinya Goto, Stefan K James, Kausik K Ray, Alexander Parkhomenko, Mikhail N Kosiborod, Darren K McGuire, Robert A Harrington, on behalf of the THEMIS Steering Committee and Investigators**

*Drs. Bhatt and Steg contributed equally to this article.

**Please see the Appendix for the THEMIS-PCI Steering Committee and Investigators.

Brigham and Women’s Hospital Heart and Vascular Center and Harvard Medical School Boston, MA, USA (Prof D L Bhatt MD); FACT (French Alliance for Cardiovascular Trials), an F-CRIN network, Département Hospitalo-Universitaire FIRE, AP-HP, Hôpital Bichat, Université Paris-Diderot, INSERM U-1148, Paris, France (Prof P G Steg MD); NHLI, Imperial College, Royal Brompton Hospital, London, United Kingdom (Prof P G Steg MD); Population Health Research Institute, Hamilton Health Sciences and McMaster University, Hamilton, Canada (S R Mehta MD); Li Ka Shing Knowledge Institute, St. Michael’s Hospital, University of Toronto, Toronto, Canada (Prof L A Leiter MD); AP-HP, Hôpital Saint Antoine, Department of Clinical Pharmacology-Clinical Research Platform (URCEST-CRB-CRCEST), and Sorbonne-Université Paris, France (T Simon MD); National Heart and Lung Institute, Imperial College and Royal Brompton Hospital London SW3 United Kingdom (K Fox MD); Department of Medical Sciences, Cardiology, Uppsala Clinical Research Center, Uppsala University, Uppsala, Sweden (C Held MD); AstraZeneca Gothenburg, R&D BioPharmaceuticals, Mölndal, Sweden (M Andersson MSc, A Himmelmann MD, W Ridderstråle MD, J Chen MD); Baim Institute for Clinical Research, Boston, MA, USA (Yang Song, MS); Department of Medicine, Estudios Clínicos Latino América, Rosario, Argentina (R Diaz MD); Department of Medicine (Cardiology), Tokai University School of Medicine, Isehara, Japan (Prof S Goto MD); Department of Medical Sciences, Cardiology and Uppsala Clinical Research Center, Uppsala University, Uppsala, Sweden (Prof S K James MD); Imperial Centre for Cardiovascular Disease Prevention, Department of Primary Care and Public Health, Imperial College, London, United Kingdom (K K Ray FRCP); Institute of Cardiology, Emergency Cardiology Department, Kiev, Ukraine (A P Parkhomenko MD); Saint Luke’s Mid-America Heart Institute, University of Missouri–Kansas City, Kansas City, MO, USA and The George Institute for Global Health, Sydney, Australia (M N Kosiborod MD); University of Texas Southwestern Medical Center, Dallas, TX, USA (D K McGuire, MD); Stanford University, Stanford, CA, USA (Prof R A Harrington MD)

123

456789

1011

12

1314151617181920212223242526272829303132333435363738394041

The THEMIS trial (NCT01991795) was sponsored by AstraZeneca.

Address for correspondence:

Deepak L. Bhatt MD, MPHBrigham and Women’s Hospital Heart & Vascular Center, Harvard Medical School 75 Francis Street, Boston, MA 02115 USAPhone: 1-857-307-1992dlbhattmd@post.harvard.edu

or

Ph. Gabriel Steg MD, Department of Cardiology, Hôpital Bichat, Paris, France, 46 rue Henri Huchard, 75018 Paris, Francegabriel.steg@aphp.fr

42

43

44

454647484950

51

525354

Summary

Background Patients with stable coronary artery disease (CAD) and diabetes with prior percutaneous coronary intervention (PCI) have been previously treated with dual antiplatelet therapy and may have a favorable benefit-risk profile with aspirin plus ticagrelor.

Methods The Effect of Ticagrelor on Health Outcomes in diabEtes Mellitus patients Intervention Study (THEMIS) randomised 19,220 patients with stable CAD and diabetes but without prior myocardial infarction or stroke to either ticagrelor or placebo (1:1 ratio) on top of low-dose aspirin. The primary efficacy outcome was cardiovascular death, myocardial infarction, or stroke. Net clinical benefit was prespecified as irreversible harm (all-cause mortality, myocardial infarction, stroke, fatal bleeding, or intracranial haemorrhage). Prior PCI, the major inclusion criterion, was a prespecified subgroup.

Findings THEMIS-PCI consisted of 11,154 patients (58% of THEMIS) with a history of prior PCI. Median follow-up was 3⋅3 years. Compared with placebo, ticagrelor reduced the primary efficacy outcome to 404/5558 (7⋅3%) from 480/5596 (8⋅6%) (HR 0⋅85, 95% CI 0⋅74–0⋅97, P=0⋅01), NNT=84; (whereas it did not in patients without PCI, pinteraction=0⋅16). Cardiovascular death was 174 (3⋅1%) with ticagrelor and 183 (3⋅3%) with placebo (HR 0⋅96, 95% CI 0⋅78–1⋅18); all-cause death was 282 (5⋅1%) and 323 (5⋅8%) (HR 0⋅88, 95% CI 0⋅75–1⋅03). TIMI Major bleeding occurred in 111 (2⋅0%) with ticagrelor and 62 (1⋅1%) with placebo (HR 2⋅03, 95% CI 1⋅48–2⋅76, P<0⋅0001), NNH=123. Fatal bleeding occurred in 6/5536 patients (0⋅1%) with ticagrelor and 6/5564 (0⋅1%) with placebo (HR 1⋅13, 95% CI 0⋅36, 3⋅50, P=0⋅83). Intracranial hemorrhage occurred in 33/5536 (0⋅6%) and 31/5564 (0⋅6%) of patients (HR 1⋅21, 95% CI 0⋅74–1⋅97, P=0⋅45). Ticagrelor improved net clinical benefit: 519/5558 (9⋅3%) versus 617/5596 (11⋅0%), HR=0⋅85, 95% CI 0⋅75–0⋅95, P=0⋅005, in contrast to patients without PCI where it did not, Pinteraction=0⋅012. Benefit was present irrespective of time from most recent PCI.

Interpretation In stable CAD patients with diabetes and prior PCI, ticagrelor added to aspirin reduced cardiovascular death, MI, and stroke, although with increased major bleeding. The net clinical benefit appeared more favourable in patients with prior PCI than in patients without prior PCI.

Funding AstraZeneca, THEMIS ClinicalTrials.gov number, NCT01991795.

55565758596061626364656667686970717273747576777879808182838485868788899091

92

93

94

Research in context

Evidence before this study

Patients with diabetes mellitus and stable coronary artery disease are at high risk of ischaemic events. Dual antiplatelet therapy is known to reduce the risk of cardiovascular events for at least 12 months after acute coronary syndromes, and longer durations of therapy provide continued benefit in patients at low bleeding risk. Whether stable patients with diabetes and a history of prior PCI but without prior myocardial infarction or stroke benefit is unknown. We searched PubMed for published randomised controlled trials of dual antiplatelet therapy in patients with diabetes mellitus. The majority of trials suggested greater degrees of efficacy in patients with diabetes versus without diabetes with dual antiplatelet therapy versus aspirin alone. However, none of those trials was designed to examine patients with diabetes specifically.

Added value of this study

The THEMIS trial investigated ticagrelor versus placebo on top of aspirin in stable patients with type 2 diabetes mellitus and angiographic evidence of coronary artery disease, but without prior myocardial infarction or stroke, and found a significant reduction in ischaemic events, though no significant net clinical benefit in the overall population when fatal or intracranial bleeding was also considered. THEMIS-PCI, the prespecified subgroup analysis of stable patients with diabetes and fulfilling the major inclusion criterion of having a history of prior percutaneous coronary intervention from THEMIS, provides the largest sample to date of such patients, with a median follow up duration of 3⋅3 years and finds a significant 15% reduction in ischaemic events with ticagrelor when added to aspirin. Furthermore, when considering irreversible harms such as fatal or intracranial bleeding in addition to effects on ischaemic events, there appears to be a favorable net clinical benefit in the patients with a history of prior PCI, though not in those without a history of prior PCI. TIMI major bleeding was increased in both groups. Those patients undergoing PCI specifically receiving stents had the most substantial risk reductions, with even more pronounced effects in those who had received drug eluting stents. The benefit was present irrespective of time from most recent percutaneous coronary intervention, though there was a suggestion of greater benefit when the history of PCI was less remote. In a post hoc landmark analysis at 1 year, there was a consistent reduction in ischemic events before and after 1 year, which was statistically significant after 1 year.

Implications of all the available evidence

In stable patients with type 2 diabetes and a history of prior percutaneous coronary intervention, in particular stent implantation, prolonged therapy with ticagrelor plus aspirin reduces ischaemic events and appears to provide significant net clinical benefit over aspirin alone, though major bleeding is increased. In patients with diabetes who are carefully selected to be at low risk of bleeding, prolonged therapy with ticagrelor and aspirin may be considered after coronary stent implantation.

95

96

979899

100101102103104105106

107

108109110111112113114115116117118119120121122123124125126127

128

129130131132133134

Introduction

Patients with a history of diabetes mellitus are at increased risk of ischaemic events, particularly those who have undergone prior percutaneous coronary intervention (PCI) with stenting. This large group of patients has a substantial future risk of coronary, cerebral, and peripheral events.1 While advances in stent design have greatly reduced the risk of stent thrombosis, the underlying coronary atherosclerosis that led to the need for stenting, especially in the patient with diabetes, portends a future risk for atherosclerotic events.2–4

In this context, antiplatelet therapy has a major role.5–7 Dual antiplatelet therapy (DAPT) with a P2Y12 inhibitor plus aspirin is the standard of care for at least 12 months in patients with acute coronary syndromes (ACS).18–15 Beyond 12 months, data support use of DAPT in ACS patients with high ischaemic risk but low bleeding risk, though identifying such patients clinically can be a challenge.16–22 For stable coronary artery disease requiring stenting, current guidelines recommend 6 months of DAPT, though potentially shorter or longer durations may be considered depending on bleeding and ischaemic risks. In particular, patients who have received stents and already tolerated a period of DAPT without bleeding complications may be best suited to continue or reinitiate therapy if they remain at high ischaemic risk, as would be the case for the majority of patients with diabetes mellitus.

The Effect of Ticagrelor on Health Outcomes in diabEtes Mellitus patients Intervention Study (THEMIS) trial demonstrated a significant 10% relative risk reduction (hazard ratio [HR] 0⋅90, P=0⋅038) in the primary endpoint of cardiovascular death, myocardial infarction, or stroke in 19,220 patients with coronary artery disease and type 2 diabetes mellitus, though the effect on the net clinical benefit endpoint did not reach statistical significance (HR 0⋅93, P=0⋅12).23,24 In this prespecified analysis of patients enroled based upon the major trial inclusion criterion, we examine those patients from THEMIS who had a history of PCI, including those specifically receiving stents, as well as time since most recent PCI, and the potential benefits of ticagrelor versus placebo on top of aspirin. We had hypothesised that this easily, clinically identifiable cohort of patients fulfilling the major inclusion criterion of the study would be the group most likely to have a favorable balance of efficacy and safety and in whom ticagrelor could serve an unmet clinical need.

Methods

Study design and participants

The details of the design of THEMIS have been previously published.23 In brief, patients ≥ 50 years of age with type 2 diabetes mellitus receiving anti-hyperglycemic medications for at least six months and with stable coronary artery disease as described

135

136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169

170

171

172173174

below were randomised in a double-blind manner to receive ticagrelor or placebo on top of aspirin.

Eligible patients were randomized in a 1:1 ratio to either ticagrelor or placebo, using an interactive voice-response or web-response system. Randomization codes were generated in blocks (of constant size) allocated to each centre, without stratification, to ensure balance between treatment groups. The treatment allocation was double blind, and patients were given ticagrelor tablets or matching ticagrelor placebo tablets, in identical appearance.

Follow-up visits were scheduled as follows: Telephone contacts were performed at 7 and 30 days after randomization. On-site visits were performed at 3, 6, 12, 18, 24, 30 months etc. after randomization, alternated by scheduled telephone contacts at 9, 15, 21, 27 months etc. after randomization. Premature Treatment Discontinuation Visits were performed for patients who prematurely and permanently discontinue study medication. A Study Closure Visit was performed within 60 days of the Primary Analysis Censoring Date.

Compliance to study drug treatment was derived from pill counts (i.e., number of pills taken divided by the expected number of pills taken. Compliance with randomised study drug was high; greater than 80% in most patients (78⋅0%) with a median of 94⋅8% compliance, and similar across the treatment groups. All stops of study medication prescribed by the investigator or treating physician were to be recorded, including any non-prescribed temporary stops (>1 week) of study medication.

Patients with prior MI or stroke were excluded. Planned use of P2Y12 inhibitors (e.g., clopidogrel, ticlopidine, prasugrel), dipyridamole, or cilostazol were exclusion criteria. In case of PCI during the study, study drug was temporarily stopped, and patients were switched to open label DAPT for the appropriate duration as determined by guidelines and the treating physician. If a patient developed an indication for anticoagulation, study drug was stopped, and patients were treated as needed with oral anticoagulation.

To qualify for enrolment, patients with type 2 diabetes mellitus also had to have one of three other mutually non-exclusive criteria: a history of previous PCI or of coronary artery bypass grafting or documentation of angiographic stenosis of ≥50% in at least one coronary artery. Patients with a history of PCI accounted for the majority (58%) of the total population (Figure S1). Prior history of PCI was a prespecified subgroup in the protocol and statistical analysis plan. This subgroup was further divided into patients who did or did not receive coronary stents. Time from most recent PCI was also prespecified as <1 year, 1–3 years, >3 years.

Additional subgroups presented include age (<65, 65-75, >75), sex, race, geographic region, body mass index, aspirin dose at baseline, HbA1c at randomisation, creatinine clearance, duration of diabetes, insulin use at baseline, history of angina, multivessel coronary artery disease, coronary artery bypass graft surgery, time since PCI, type of stent, proton pump inhibitor use at baseline, current smoking, and history of polyvascular disease.

175176

177178179180181182

183184185186187188189

190191192193194195196197198199200201202

203204205206207208209210

211212213214215216

The trial initially started with 90 mg twice daily as the dose of ticagrelor. Approximately a year after the initiation of THEMIS, the Prevention of Cardiovascular Events in Patients with Prior Heart Attack Using Ticagrelor Compared to Placebo on a Background of Aspirin (PEGASUS) study showed that the efficacy profiles of ticagrelor 90 mg twice daily and 60 mg twice daily (administered with low-dose aspirin) were similar, and there was evidence that the lower dose had a better tolerability profile. Therefore, the ticagrelor dose in THEMIS was switched to 60 mg twice daily.19

Outcomes

The primary efficacy endpoint was time to first event of the composite of cardiovascular death, MI, or stroke in the intention-to-treat population.23,24. Safety analyses were performed on-treatment, patients who took at least one dose of study drug were analyzed, and events that occurred between randomisation and up to and including seven days post last dose were included in the analyses. The primary safety endpoint was Thrombolysis in Myocardial Infarction (TIMI) Major bleeding in the on-treatment population. Net clinical benefit was prespecified as irreversible harms evaluated as time to first event of the composite of all-cause mortality, MI, stroke, fatal bleeding, or intracranial haemorrhage in the intention-to-treat population. A more expansive prespecified endpoint incorporating coronary, cerebral, and peripheral ischaemic events was determined. Endpoints presented in this paper were adjudicated by a clinical endpoints committee masked to treatment allocation.

Statistical analysis

To analyze time-to-event variables, a Cox proportional hazards model with treatment group as a covariate and the Efron method for ties were used. The proportional hazards assumption was verified visually for the entire trial, using a log-cumulative hazard plot for the primary efficacy endpoint, which showed parallel and overlaid lines. The Wald statistic was used to calculate confidence intervals and P-values. P-values for interaction between treatment group and subgroup were calculated from the Cox proportional hazards model with treatment group, subgroup, and interaction term as explanatory variables. For efficacy analyses according to the intention-to-treat-principle, all randomised patients were analyzed and events that occurred prior to or at the primary analysis censoring date were included in the analysis. Event-free patients were censored at the earlier of the primary analysis censoring date or date of last clinical event assessment. Safety analyses were performed on-treatment, patients who took at least one dose of study drug were analyzed, and events that occurred between randomisation and up to and including seven days post last dose were included in the analyses. P-values for trend were estimated with the log-rank trend test comparing strata survival curves. No adjustments were made for multiple comparisons, and therefore the P-values presented are nominal.

Specific analyses were prespecified to test the consistency of treatment effect for the two doses of ticagrelor. Primary efficacy and safety analyses were repeated in the subgroup of patients randomized after the implementation of the dose reduction, a time

217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255256257258259260261262

dependent Cox model with a factor for treatment group and a time-dependent variable for time on ticagrelor 60 mg was used to estimate the treatment effect of ticagrelor 60 mg vs placebo, and on-treatment analyses for each dose were performed.

Role of the funding sourceThe trial was designed and governed by an academic Executive Committee, which also included non-voting members representing the sponsor, AstraZeneca. AstraZeneca funded the trial and was involved with the data collection and analysis. Baim Clinical Research Institute, an academic research organisation, independently validated all the data contained in this manuscript. The authors made the decision to publish the paper. Dr. Deepak L. Bhatt and Dr. Ph. Gabriel Steg had full access to the data and final responsibility for the decision to submit for publication.

Results

Baseline characteristics

THEMIS-PCI was comprised of 11,154 patients (58% of the overall THEMIS trial) with a history of prior PCI, enroled at 1315 sites in 42 countries. The first patient was enrolled on 10 February 2014, the last on 24 May 2016, and the last patient visit was on 25 January 2019. The primary analysis censoring date for efficacy analyses was October 29, 2018. The median follow-up was 3⋅3 years (maximum 4⋅8 years up to the primary analysis censoring date). As expected, there were significant differences in baseline characteristics between the PCI and no PCI groups (Table S1). The baseline characteristics of the patients were well matched in the ticagrelor and placebo groups (Table S2). The median duration of diabetes was 10⋅0 years. Patients were on a median of 2⋅0 antihyperglycemic medications. The median time since the most recent PCI was 3⋅3 years (interquartile range: 1⋅5–6⋅6 years).

Efficacy endpoints

Compared with placebo, ticagrelor reduced the primary efficacy outcome to 404/5558 (7⋅3%) from 480/5596 (8⋅6%) (HR, 0⋅85; 95% confidence interval [CI] 0⋅74–0⋅97; P=0⋅01) (Figure 1, top left panel) in the THEMIS-PCI population (number needed to treat [NNT]=84), whereas in the 8066 (42⋅0% of the overall THEMIS trial) patients with no history of PCI, there was no apparent benefit (8⋅2% vs 8⋅4%, HR, 0⋅98; 95% CI 0⋅84–1⋅14; P=0⋅76), Pinteraction=0⋅16 (Figure 1, top right panel). The benefit on the primary efficacy outcome was consistent in the examined subgroups, with no significant interactions (Figure S2). In THEMIS-PCI, ticagrelor reduced MI, including ST-segment elevation MI, and stroke (Table 1). There was a trend towards a lower rate of MI related to stent thrombosis in THEMIS-PCI (P=0⋅072). Among those with a history of PCI, the occurrence of STEMI was 16/5558 (0⋅3%) for ticagrelor and 51/5596 (0⋅9%) for placebo (HR 0⋅32, 95% CI 0⋅18–0⋅55, P<0⋅0001). Definite stent thrombosis was 8/5558 (0⋅1%) for ticagrelor and 14/5596 (0⋅3%) for placebo (HR 0⋅58, 95% CI 0⋅24–1⋅37, P=0⋅21) and

263264265266267268269270271272273274275

276

277

278279280281282283284285286287288289290291292293294295296297298299300301302303304

definite or probable stent thrombosis was 9/5558 (0⋅2%) with ticagrelor, 18/5596 (0⋅3%) with placebo (HR.0⋅50, 95% CI 0⋅23–1⋅12, P=0⋅094).

A more expansive prespecified endpoint incorporating coronary, cerebral, and peripheral ischaemic events was also significantly reduced with ticagrelor versus placebo in THEMIS-PCI. The composite of all-cause death, MI, stroke, acute limb ischaemia, or major amputation of vascular etiology was reduced to 9⋅0% from 11⋅0% in those with a history of PCI (HR 0⋅82, 95% CI 0⋅72–0⋅92, P=0⋅0007), whereas there was no apparent benefit in those without a history of PCI (HR 1⋅00, 95% CI 0⋅88–1⋅15, P=0⋅97), Pinteraction=0⋅023.Bleeding

TIMI and PLATelet inhibition and patient outcomes (PLATO) major bleeding, as well as major or minor bleeding, were significantly increased with ticagrelor, both in the PCI and no PCI groups (Table 2). In the PCI group, TIMI major bleeding was increased to 2⋅0% from 1⋅1%, (HR 2⋅03, 95% CI 1⋅48–2⋅76, P<0⋅0001), number needed to harm [NNH]=123; in the no PCI group, TIMI major bleeding was increased to 2⋅4% from 1⋅0% (HR 2⋅79, 95% CI 1⋅91–4⋅06, P<0⋅0001), Pinteraction=0⋅20. TIMI major bleeding over time in patients with or without a history of PCI is shown in Figure 1, bottom left and right panels, respectively. The increase in TIMI major bleeding was consistent across subgroups (Figure S3).

BARC type 2, 3, 4, or 5 bleeding was also significantly increased in both the PCI and no PCI groups. However, in this case the interaction term was significant, suggesting a higher bleeding risk in the no PCI versus PCI groups.

Fatal bleeding occurred in 6/5536 patients (0⋅1%) in the ticagrelor group and 6/5564 (0⋅1%) in the placebo group (HR 1⋅13, 95% CI 0⋅36–3⋅50, P=0⋅83). Intracranial haemorrhage occurred in 33/5536 (0⋅6%) and 31/5564 (0⋅6%) of patients (HR 1⋅21, 95% CI 0⋅74–1⋅97, P=0⋅45), whereas in the no prior PCI patients, rates were 0⋅9% versus 0⋅4%, (HR 2⋅74, 95% CI, 1⋅51–5⋅00, P=0⋅001), Pinteraction=0⋅036.

Net clinical benefit

Ticagrelor significantly improved the net clinical benefit of irreversible harms: 519/5558 (9⋅3%) vs 617/5596 (11⋅0%), HR=0⋅85, 95% CI 0⋅75–0⋅95, P=0⋅005 (Table S3, Figure 2), whereas, in patients with no history of PCI, there was no net benefit (11⋅1% versus 10⋅5%, HR 1⋅06, 95% CI 0⋅93–1⋅21, P=0⋅39), Pinteraction=0⋅012.

When net clinical benefit was redefined post hoc to encompass the primary efficacy variable (CV death, myocardial infarction or stroke) and the primary safety variable (TIMI major bleeding),then there was no significant benefit in either group: 510/5558 (9⋅2%) for ticagrelor, 529/5596 (9⋅5%) for placebo (HR 0⋅97, 95% CI 0⋅86–1⋅10, P=0⋅68) in those with a history of PCI; in those without a history of PCI, 418/4061 (10⋅3%) for

305306307308309310311312313314315316317318319320321322323324325326327328329330331332333334335336337338339340341342343344345346347348

ticagrelor, 370/4005 (9⋅2%) for placebo (HR 1⋅13, 95% CI 0⋅98–1⋅30, P=0⋅08), Pinteraction=0⋅12.

When net clinical benefit was redefined post hoc as the composite of all-cause mortality, STEMI, stroke, fatal bleeding, or intracranial haemorrhage, there was significant benefit in patients with a history of prior PCI 391/5558 (7⋅0%) for ticagrelor, 472/5596 (8⋅4%) for placebo (HR 0⋅83, 95% CI 0⋅73–0⋅95, P=0⋅008), but not in those without a history of prior PCI, ticagrelor 382/4061 (9⋅4%) for ticagrelor, 350/4005 (8⋅7%) for placebo (HR 1⋅09, 95% CI 0⋅94–1⋅26, P=0⋅26), Pinteraction=0⋅008.

Time since PCI

There was similar efficacy for the primary outcome with more recent PCI versus more remote PCI, with non-significant Pinteraction and Ptrend (HR for PCI <1 year: 0⋅54; HR for PCI 1–3 years prior: 0⋅84; HR for PCI >3 years prior: 0⋅92; Pinteraction=0⋅07; Ptrend=0⋅63). In a post hoc analysis of the primary efficacy endpoint and its components, time from most recent PCI was analyzed as a continuous variable, and there was still no significant interaction between time since PCI and treatment group, with an upper bound of the 95% CI of the HR that remained under 1 until approximately 6 years after the most recent PCI; the P-value for interaction between treatment group and time since PCI was 0⋅11 (Figure 3).

There were in total 84 patients randomized within 1 month from most recent PCI; among these there were 4/44 (9⋅1%) in the ticagrelor group and 1/40 (2⋅5%) with primary efficacy endpoint events (Table S4). Corresponding numbers for patients who were randomized within 6 months were 15/229 (6⋅6%) in the ticagrelor group and 27/223 (12⋅1%) in the placebo group with primary efficacy endpoint events, HR 0⋅53, 95% CI 0⋅28 –1⋅00, P=0⋅0489. Those rates in patients randomized after 6 months were 389/5328 (7⋅3%) in the ticagrelor group and 453/5372 (8⋅4%) in the placebo group, HR 0⋅87 95% CI 0⋅76–0⋅99 P=0⋅0359, Pinteraction=0⋅13.

On a post hoc landmark analysis of the primary efficacy endpoint (Figure 4), up to 1 year, there were 123/5558 (2⋅2%) patients with events for ticagrelor and 140/5596 (2⋅5%) patients with events for placebo, HR 0⋅89, 95% CI 0⋅70–1⋅13, P=0⋅33. For the analysis of 1 year and onwards, there were 281/5347 (5⋅3%) patients with events for ticagrelor and 340/5385 (6⋅3%) patients with events for placebo, HR 0⋅83, 95% CI 0⋅71–0⋅97, P=0⋅021. Events for patients with and with and without a history of PCI are shown in Table S5.

Patients with a history of stenting

In the 10,295 PCI patients (92% of THEMIS-PCI) who specifically had a history of coronary stenting, the primary efficacy endpoint was reduced with ticagrelor to 367/5101 (7⋅2%) from 457/5194 (8⋅8%) (HR 0⋅81, 95% CI 0⋅71–0⋅93, P=0⋅003) (Figure S4). In these patients with a history of stenting at any time, there was also no significant trend

349350351352353354355356357358359360361362363364365366367368369370371372373374375376377378379380381382383384385386387388389390391392393

towards differential benefit based on time from the last PCI procedure (HR for PCI <1 year: 0⋅58; HR for PCI 1–3 years prior: 0⋅82; HR for PCI >3 years prior: 0⋅86; Ptrend=0⋅37; Pinteraction=0⋅28). The more expansive composite of all-cause death, MI, stroke, acute limb ischaemia, or major amputation of vascular etiology was reduced to 8⋅9% from 11⋅2% in those with a history of stenting (HR 0⋅79, 95% CI 0⋅70–0⋅90, P=0⋅0002). The individual components of all-cause death, MI, and stroke were all significantly reduced (Table S6).In a prespecified analysis conducted in patients with prior drug eluting stents, the primary efficacy endpoint occurred in 233/3371 (6⋅9%) of ticagrelor patients and 297/3437 (8⋅6%) of placebo patients (HR 0⋅79, 95% CI 0⋅67–0⋅94, P=0⋅008) (Figure S5). The endpoint of irreversible harm occurred in 306/3371 (9⋅1%) of ticagrelor patients and 370/3437 (10⋅8%) placebo patients (HR 0⋅84, 95% CI 0⋅72–0⋅97 P= 0⋅021). Additional endpoints in this population are shown in Table S7.

In the patients without a history of PCI, whether treated with prior CABG or medically, there was no significant benefit of ticagrelor over placebo (Table S8)

Sensitivity analyses based on dose of ticagrelor

Prespecified sensitivity analyses showed consistency between the treatment effect of ticagrelor 60 mg twice daily and ticagrelor 90 mg twice daily and between 60 mg twice daily and the overall population for the primary efficacy endpoint (Tables S9 and S10, respectively) and for TIMI major bleeding between ticagrelor 60 mg daily and the overall population (Table S11), all in patients with a history of PCI.

Drug discontinuation and adverse events

For the THEMIS-PCI population, permanent discontinuation rates were 34⋅8% with ticagrelor and 25⋅7% with placebo, P<0⋅0001. This corresponds to an average rate of 11⋅4% and 8⋅2% per year, respectively, though there was a higher rate of discontinuation during the first year. The most common adverse events leading to discontinuation of study drug were dyspnoea (7⋅3% versus 0⋅8%, P<0⋅0001) and bleeding (4⋅7% versus 1⋅3%, P<0⋅0001) for ticagrelor versus placebo. The rates of dyspnoea leading to treatment discontinuation are shown in Figure S6.

Adding events to the ticagrelor subjects with incomplete follow-up, assuming the same event rate as in the placebo group, the result was still significant.

On-treatment analyses

In on-treatment analyses, the rates of the primary efficacy endpoint were 4⋅1% in ticagrelor group and 6⋅2% in the placebo group (HR 0⋅73, 95% CI 0⋅62–0⋅87, P=0⋅0003) (Table S12). There was also a significant reduction in all-cause mortality in the on-treatment analyses to 77/5536 (1⋅4%) from 118/5564 (2⋅1%), HR 0⋅74, 95% CI 0⋅56–

394395396397398399400401402403404405406407408409410411412413414415416417418419420421422423424425426427428429430431432433434435436437

0⋅99, P=0⋅044) in the prior PCI group, while in the no prior PCI group there was no reduction, Pinteraction=0⋅021.

Discussion

This prespecified analysis of THEMIS suggests that patients with stable CAD and type 2 diabetes mellitus who have a history of prior PCI derive significant benefit from adding ticagrelor to a regimen of aspirin. In particular, those with a history of prior PCI utilizing drug eluting stents, which is currently the predominant mode of percutaneous coronary revascularisation, appeared to derive the largest benefit compared with the rest of the THEMIS population. Post hoc landmark analyses suggested that the benefits were consistent over time but became significant one year after randomization. These data, taken together with prior data from the PLATO and PEGASUS trials, suggest that DAPT with aspirin and ticagrelor is efficacious in patients with recent acute coronary syndromes, prior myocardial infarction, and now, stable CAD and diabetes mellitus who have undergone prior stenting.12,13,19–21,24

While the overall THEMIS trial did meet its primary efficacy endpoint, there was an increase in intracranial haemorrhage (driven by traumatic intracranial haemorrhage) with ticagrelor versus placebo when added to aspirin.24 Thus, the overall net clinical benefit when examining irreversible endpoints was not significantly influenced. However, in THEMIS PCI, focusing on a large prespecified subgroup of patients fulfilling the major THEMIS inclusion criterion of having a history of prior PCI, there appeared to be a significantly favorable effect on the primary efficacy endpoint, and greater net clinical benefit (Figure 5). This important cohort of patients from THEMIS who appeared to benefit is easy to identify and, building upon prior research, is a logical group in whom to consider extended duration DAPT. Of course, a patient’s individual risks of future ischaemic events and bleeding do need to be weighed carefully, as TIMI major bleeding was significantly increased in the overall THEMIS trial,24 and in THEMIS-PCI. Also, when net clinical benefit was defined post hoc to include TIMI major bleeding, there was no longer significant benefit in the prior PCI patients. Finally, in THEMIS, stable CAD patients without a history of PCI did not derive benefit from the addition of ticagrelor to aspirin, though were still subject to at least the same or greater bleeding risks, so DAPT should generally not be used in those patients.

Overall, the subgroup with a history of PCI appeared to fare better with ticagrelor both in terms of efficacy and safety than patients without history of PCI, although interaction P values were only statistically significant for the prespecified irreversible harm net clinical benefit and for intracranial bleeding and BARC bleeding, though not for the primary efficacy endpoint or TIMI bleeding. Thus, from a statistical perspective, these subgroup findings should be interpreted with caution.

Bleeding is a concern with all antiplatelet agents, including potent agents such as ticagrelor. Potentially, use of proton pump inhibitors in patients at risk of gastrointestinal bleeding could help reduce the most common form of major bleeding. Randomised data

438439440441442

443444445446447448449450451452453454455456457458459460461462463464465466467468469470471472473474475476477478479480481482

from the Clopidogrel and the Optimization of Gastrointestinal Events Trial (COGENT) trial have demonstrated a large reduction in bleeding from DAPT in patients assigned to a proton pump inhibitor.25 An agent to reverse ticagrelor is in development and may be useful in cases of serious bleeding, such as intracranial haemorrhage.26 Risk scores may help determine bleeding risk quantitatively.22 Perhaps the best gauge of a patient’s bleeding risk is whether they pass a “bleeding stress test”2 of having been on DAPT without sustaining a bleeding complication.17 This latter point is supported by the present analysis of THEMIS which suggests greater net clinical benefit in the patients with prior PCI than in the rest of the trial population; presumably, patients with prior stenting have received a period of DAPT and would not have been enrolled in THEMIS if they had experienced a bleeding event while on DAPT. Previous studies of antithrombotics have demonstrated that history of prior PCI (versus no prior PCI) identifies a subgroup of patients with a lower risk of bleeding.27 Although there was a significant excess in TIMI major bleeding, it is reassuring that the patients in THEMIS-PCI had no significant increase in fatal or intracranial bleeding in terms of considering extended duration ticagrelor.

Despite some controversy in terms of the optimal revascularisation strategy in patients with stable coronary artery disease and diabetes mellitus, PCI especially with drug eluting stents remains an important modality for relief of angina, especially when medical therapy is insufficient for symptom control, and the extent of coronary atherosclerosis and complexity of coronary anatomy are not substantial enough to warrant coronary artery bypass surgery.28,29 As demonstrated in THEMIS-PCI, adding ticagrelor to aspirin in these stable patients who have undergone PCI in the past could potentially further improve their outcomes by reducing the risk of myocardial infarction, including ST-segment elevation myocardial infarction, stroke, and peripheral ischaemic events. Additionally, while not directly studied in THEMIS-PCI, patients with stable coronary artery disease and diabetes mellitus who have undergone PCI, are already receiving DAPT, and tolerating it well might benefit from continuing the therapy long-term. Data from the Cardiovascular Outcomes for People Using Anticoagulation Strategies (COMPASS) trial also support the concept of intensifying the anti-thrombotic regimen beyond aspirin alone (with low dose rivaroxaban) in appropriately risk-stratified patients with stable atherosclerosis who are at high ischemic risk but low bleeding risk.30,31

In contrast to observations in PLATO, where there was lower all-cause mortality with ticagrelor compared with clopidogrel in ACS patients, and in COMPASS, where there was a lower all-cause mortality with low-dose rivaroxaban compared with placebo in high risk but stable CAD patients, THEMIS-PCI did not show a reduction in all-cause mortality in the intent-to-treat analysis, although there was a lower all-cause mortality in patients with a history of prior stenting (HR 0⋅84, 95% CI 0⋅71–1⋅00, P=0⋅046), Pinteraction =0⋅016. Thus, there remains uncertainty as to the best antithrombotic approach beyond aspirin in stable patients with coronary artery disease, and further research is necessary, including head-to-head trials of DAPT and double antithrombotic therapy with an antiplatelet plus a reduced dose non-vitamin K oral anticoagulant. Potentially, patients with stent-based elevated ischemic risks would be best served with prolonged

483484485486487488489490491492493494495496497498499500501502503504505506507508509510511512513514515516517518519520521522523524525526527528

DAPT, while patients with high ischemic risk not specifically related to stents may be best suited for therapy with aspirin and low dose rivaroxaban – again, assuming that there is no history of prior bleeding and no evidence of heightened future bleeding risk, perhaps as assessed by a bleeding risk score such as PRECISE-DAPT.2,22 In the future, platelet function testing or other more global biomarker assessments of prothrombotic tendencies may help individualize therapy and guide which patients could benefit from antithrombotic intensification or de-escalation.

Limitations include that this is a subgroup analysis, (albeit a prespecified and clinically important subgroup consisting of patients who fulfilled the major inclusion criterion of an overall positive trial), and therefore, the results should be interpreted with caution, especially since the interaction P-value for the primary efficacy endpoint for the PCI versus no PCI subgroups was not statistically significant. In addition, while ticagrelor reduced the incidence of STEMI, a caveat is that information on the type of MI (STEMI vs NSTEMI), while adjudicated, could not be classified in 23⋅8% of patients. Despite this being the largest randomised study of patients with diabetes mellitus having a history of PCI performed to date, statistical power was still limited to examine the time dependency of benefit, though data from THEMIS-PCI do show efficacy of very protracted durations of DAPT, with no statistically significant difference in the magnitude of benefit as a function of the time from the most recent PCI procedure. There were patients in the prior PCI group who did not receive stents, and they did not appear to benefit from ticagrelor added to aspirin; it is unclear why, but presumably these were stable patients with lesions in coronary arteries deemed too small to stent and might just reflect lower atherosclerotic burden than in patients who received stenting of larger diameter vessels with obstructive lesions. However, detailed angiographic data were not available in THEMIS-PCI. Reassuringly, there were only a small number of patients randomized within 1 month of most recent PCI. Sensitivity analyses demonstrated that the primary efficacy results were robust and remained significant even after removing patients at 1, 3, or 6 months from most recent PCI. Finally, patients enroled in THEMIS-PCI may not reflect the highest ischaemic risk patients who could benefit the most, as doctors may already choose to treat such patients with open-label generic DAPT based on data such as the Clopidogrel for High Atherothrombotic Risk and Ischemic Stabilization, Management, and Avoidance (CHARISMA) secondary prevention subgroup analysis.16,17

In conclusion, in the large subgroup of patients fulfilling the major inclusion criteria of history of prior PCI in THEMIS, ticagrelor added to aspirin appeared to significantly reduce cardiovascular death, MI, or stroke, and importantly, also appeared to provide a favorable net clinical benefit in this easily identified cohort. In patients with diabetes mellitus undergoing PCI who are carefully assessed to be at low bleeding risk and high ischemic risk and have tolerated antiplatelet therapy, long term DAPT may be considered.

529530531532533534535536537538539540541542543544545546547548549550551552553554555556557558559560561562563564565566567568569570

Authors’ Contributions

Dr. Deepak L. Bhatt and Dr. Ph. Gabriel Steg contributed equally. They designed the study, analyzed and interpreted the data, and wrote the first draft of this paper.Marielle Andersson and Yang Song analyzed the data. All authors were involved with data interpretation and reviewed and revised the draft.

Role of the funding source

The trial was designed and governed by an academic Executive Committee, which also included non-voting members representing the sponsor, AstraZeneca. AstraZeneca funded the trial and was involved with the data collection and analysis. Baim Clinical Research Institute, an academic research organisation, independently validated all the data contained in this manuscript. The authors made the decision to publish the paper. Dr. Deepak L. Bhatt and Dr. Gabriel Steg had full access to the data and final responsibility for the decision to submit for publication.

Declaration of Interests

DLB reports grants from AstraZeneca, during the conduct of the study; grants from Amarin, grants from AstraZeneca, grants from Bristol-Myers Squibb, grants from Eisai, grants from Ethicon, grants from Medtronic, grants from sanofi aventis, grants from The Medicines Company, other from FlowCo, other from PLx Pharma, other from Takeda, personal fees from Duke Clinical Research Institute, personal fees from Mayo Clinic, personal fees from Population Health Research Institute, personal fees, non-financial support and other from American College of Cardiology, personal fees from Belvoir Publications, personal fees from Slack Publications, personal fees from WebMD, personal fees from Elsevier, other from Medscape Cardiology, other from Regado Biosciences, other from Boston VA Research Institute, personal fees and non-financial support from Society of Cardiovascular Patient Care, non-financial support from American Heart Association, personal fees from HMP Global, grants from Roche, personal fees from Harvard Clinical Research Institute (now Baim Institute for Clinical Research), other from Clinical Cardiology, personal fees from Journal of the American College of Cardiology, other from VA, grants from Pfizer, grants from Forest Laboratories/AstraZeneca, grants from Ischemix, other from St. Jude Medical (now Abbott), other from Biotronik, other from Cardax, other from Boston Scientific, grants from Amgen, grants from Lilly, grants from Chiesi, grants from Ironwood, personal fees from Cleveland Clinic, personal fees from Mount Sinai School of Medicine, other from Merck, grants from Abbott, grants from Regeneron, other from Svelte, grants and other from PhaseBio, grants from Idorsia, grants from Synaptic, personal fees from TobeSoft, personal fees and other from Boehringer Ingelheim, personal fees from Bayer, other from Novo Nordisk, other from Fractyl, personal fees from Medtelligence/ReachMD, personal fees from CSL Behring, other from Cereno Scientific, outside the submitted work.

571

572573574575576577

578579580581582583584

585586

587588589590591592593594595596597598599600601602603604605606607608609610611612613

PGS reports personal fees and non-financial support from AstraZeneca, during the conduct of the study; grants and personal fees from Bayer/Janssen, grants and personal fees from Merck, grants and personal fees from Sanofi, grants and personal fees from Amarin, personal fees from Amgen, personal fees from Bristol Myers Squibb, personal fees from Boehringer-Ingelheim, personal fees from Pfizer, personal fees from Novartis, personal fees from Regeneron, personal fees from Lilly, personal fees from AstraZeneca, grants and personal fees from Servier, personal fees from Novo Nordisk, personal fees from Idorsia, outside the submitted work.

SRM discloses the following relationships: Research grants to Population Health Research Institute from AstraZeneca, Abbott, Boston Scientific and Sanofi. Consultant to AstraZeneca, Bayer, Biosensors and Sanofi. LAL discloses the following relationships- personal fees from AstraZeneca, during the conduct of the study; grants and personal fees from AstraZeneca, grants and personal fees from Boehringer Ingelheim, grants and personal fees from Eli Lilly, grants and personal fees from Janssen, personal fees from Merck, grants and personal fees from Novo Nordisk, grants and personal fees from Sanofi, personal fees from Servier, grants from Esperion, grants from GSK, grants from Kowa, grants from The Medicines Company.

TS discloses personal fees from Astrazeneca, during the conduct of the study; grants from Astrazeneca, Daiichi-Sankyo, Eli-Lilly, GSK, MSD, Novartis, Sanofi, personal fees from BMS, personal fees from Novartis, from Sanofi.KF reports personal fees from AstraZeneca, during the conduct of the study; personal fees and non-financial support from Servier, personal fees from Taurx, personal fees and non-financial support from Broadview Ventures, personal fees from CellAegis, personal fees from Celixir; and Director of Vesalius Trials Ltd.CH discloses grants and personal fees from AstraZeneca, and personal fees from Bayer and Idorsia.

MA, AH, WR and JC are employees of AstraZeneca.

YS is an employee of Baim Clinical Research Institute.

RD reports grants from the TIMI GROUP, grants from MHICC, grants from LEPETIT, grants from SANOFI, grants from LEPETIT, grants from DALCOR, grants from DCRI,SG reports personal fees from AstraZeneca, grants from MEXT/JSPS KAKENHI 17K19669, partly by 18H01726 and 19H03661, grants from The Vehicle Racing Commemorative Foundation, grants from Nakatani Foundation for Advancement of Measuring Technologies in Biomedical Engineering, grants from Bristol-Myers Squibb from their independent research support project (33999603), grants from Sanofi, grants from Pfizer, grants from Ono, personal fees from The American Heart Association.SKJ discloses grants from AstraZeneca, Jansen, Bayer, PhaseBio, and The Medicines Company.

614615616617618619620621622623624625626627628629630631632633634635636637638639640641642643644645646647648649650651652653654655656657658

KKR reports personal fees from Astra Zeneca, during the conduct of the study; personal fees from Abbvie, grants and personal fees from Amgen, personal fees from Astra Zeneca, grants and personal fees from Sanofi, grants and personal fees from Regeneron, grants and personal fees from MSD, grants and personal fees from Pfizer, personal fees from The Medicines Company, personal fees from Resverlogix, personal fees from Akcea, personal fees from Boehringer Ingelheim, personal fees from Novo Nordisk, personal fees from Takeda, personal fees from Kowa, personal fees from Algorithm, personal fees from Cipla, personal fees from Cerenis, personal fees from Dr Reddy’s, personal fees from Lilly, personal fees from Zuellig Pharma, support from the Imperial BRC, outside the submitted work; Medical Science, section Editor for Thrombosis and Hemostasis.

AP reports grants and personal fees from Astra Zeneca, during the conduct of the study; grants from Amgen, grants from Bayer, grants from BMS, grants from CSL Behring, personal fees from Servier, grants and personal fees from Sanofi.

MNK reports grants, personal fees and other from AstraZeneca, during the conduct of the study; grants, personal fees and other from AstraZeneca, grants and personal fees from Boehringer Ingelheim, personal fees from Sanofi, personal fees from Amgen, personal fees from NovoNordisk, personal fees from Merck (Diabetes), personal fees from Eisai, personal fees from Janssen, personal fees from Bayer, personal fees from GlaxoSmithKline, personal fees from Glytec, personal fees from Intarcia, personal fees from Novartis, personal fees from Applied Therapeutics, personal fees from Amarin.

DKM reports the following relationships: honoraria for clinical trials leadership from GlaxoSmithKline, Janssen, Lexicon, AstraZeneca, Sanofi Aventis, Boehringer Ingelheim, Merck & Co, Pfizer, Novo Nordisk, Eisai Inc., Esperion, Lilly USA; and honoraria for consultancy from AstraZeneca, Lilly USA, Boehringer Ingelheim, Merck & Co, Novo Nordisk, Metavant, Applied Therapeutics, Sanofi Aventis

RAH reports grants from Astra Zeneca, during the conduct of the study; grants from Sanofi Aventis, grants from Portola, grants from Janssen, grants from BMS, grants from Novartis, grants from The Medicines Company, grants from CSL, outside the submitted work; and Consulting (none after July 1, 2018): Amgen, Gilead, MyoKardia.

Data sharing

Please contact the corresponding authors with any proposals for analysis.

Acknowledgments

The THEMIS study was funded and sponsored by AstraZeneca R&D. Details of the THEMIS organisation and a complete list of the investigators are provided in the appendix. The statistical analyses were done by AstraZeneca and validated by independent statisticians at the Baim Institute for Clinical Research, Boston, MA, USA, with funding from AstraZeneca. This was in part supported by the RHU iVASC grant

659660661662663664665666667668669670671672673674675676677678679680681682683684685686687688689690691692693694695696697698699700701702703704

‘#ANR-16-RHUS-00010’ from the French National Research Agency (ANR) as part of the “Investissements d’Avenir” program. The authors had full access to all data. We would like to thank C. Michael Gibson, MS, MD (Cardiovascular Division, Department of Medicine, Beth Israel Deaconess Medical Center, Boston, MA and Baim Institute for Clinical Research, Boston, MA) and Gheorghe Doros, PhD (Baim Clinical Research Institute, Boston, MA, Department of Biostatistics, Boston University, Boston, MA) for supervising the statistical analyses. We would like to thank Jayne Prats PhD for assistance with editing (limited to collation of author comments and formatting for submission), with funding from AstraZeneca. We also thank the trial participants, investigators, and site staff who were involved in the conduct of the trial.

705706707708709710711712713714

715

References

1 Bhatt DL, Eagle KA, Ohman EM et al. Comparative determinants of 4-year cardiovascular event rates in stable outpatients at risk of or with atherothrombosis. JAMA 2010; 304: 1350–7.

2 Bavry AA, Bhatt DL. Appropriate use of drug-eluting stents: balancing the reduction in restenosis with the concern of late thrombosis. Lancet 2008; 371: 2134–43.

3 Sabate M, Cequier A, Iñiguez A, et al. Everolimus-eluting stent versus bare-metal stent in ST-segment elevation myocardial infarction (EXAMINATION): 1 year results of a randomised controlled trial. Lancet 2012; 380: 1482–90.

4 Bhatt DL. EXAMINATION of new drug-eluting stents--top of the class! Lancet 2012; 380: 1453–5.

5 Bhatt DL, Marso SP, Hirsch AT, Ringleb PA, Hacke W, Topol EJ. Amplified benefit of clopidogrel versus aspirin in patients with diabetes mellitus. Am J Cardiol 2002; 90: 625–8.

6 Wiviott SD, Braunwald E, Angiolillo DJ, et al. Greater clinical benefit of more intensive oral antiplatelet therapy with prasugrel in patients with diabetes mellitus in the trial to assess improvement in therapeutic outcomes by optimizing platelet inhibition with prasugrel thrombolysis in myocardial infarction 38. Circulation 2008; 118: 1626–36.

7 James S, Angiolillo DJ, Cornel JH, et al. Ticagrelor vs. clopidogrel in patients with acute coronary syndromes and diabetes: a substudy from the PLATelet inhibition and patient outcomes (PLATO) trial. Eur Heart J 2010; 31: 3006–16.

8 The Clopidogrel in Unstable Angina to Prevent Recurrent Events Trial Investigators. Effects of clopidogrel in addition to aspirin in patients with acute coronary syndromes without ST-segment elevation. N Engl J Med 2001; 345: 494–502.

9 Mehta SR, Yusuf S, Peters RJG, et al. Effects of pretreatment with clopidogrel and aspirin followed by long-term therapy in patients undergoing percutaneous coronary intervention: the PCI-CURE study. Lancet 2001; 358: 527–33.

10 CURRENT-OASIS 7 Investigators. Dose comparisons of clopidogrel and aspirin in acute coronary syndromes. N Engl J Med 2010; 363: 930–42.

11 Mehta SR, Tanguay JF, Eikelboom JW, et al. Double-dose versus standard-dose clopidogrel and high-dose versus low-dose aspirin in individuals undergoing percutaneous coronary intervention for acute coronary syndromes (CURRENT-OASIS 7): a randomised factorial trial. Lancet 2010; 376: 1233–43.

12 Wallentin L, Becker RC, Budaj A, et al. Ticagrelor versus clopidogrel in patients with acute coronary syndromes. N Engl J Med 2009; 361: 1045–57.

716

717718719

720721

722723724

725726

727728729

730731732733734

735736737

738739740

741742743

744745

746747748749

750751

13 Cannon CP, Harrington RA, James S, et al. Comparison of ticagrelor with clopidogrel in patients with a planned invasive strategy for acute coronary syndromes (PLATO): a randomised double-blind study. Lancet 2010; 375: 283–93.

14 Hahn JY, Song YB, Oh JH et al. 6-month versus 12-month or longer dual antiplatelet therapy after percutaneous coronary intervention in patients with acute coronary syndrome (SMART-DATE): a randomised, open-label, non-inferiority trial. Lancet 2018; 391: 1274–84. 

15 Motovska Z, Bhatt DL. 12 months of DAPT after acute coronary syndrome still beats 6 months. Lancet 2018; 391: 1240–2.

16 Bhatt DL, Fox KA, Hacke W, et al. Clopidogrel and aspirin versus aspirin alone for the prevention of atherothrombotic events. N Engl J Med 2006; 354: 1706–17.

17 Bhatt DL, Flather MD, Hacke W, et al. Patients with prior myocardial infarction, stroke, or symptomatic peripheral arterial disease in the CHARISMA trial. J Am Coll Cardiol 2007; 49: 1982–8.

18 Wiviott SD, White HD, Ohman EM, et al. Prasugrel versus clopidogrel for patients with unstable angina or non-ST-segment elevation myocardial infarction with or without angiography: a secondary, prespecified analysis of the TRILOGY ACS trial. Lancet 2013; 382: 605–13. 

19 Bonaca MP, Bhatt DL, Cohen M, et al. Long-term use of ticagrelor in patients with prior myocardial infarction. N Engl J Med 2015; 372: 1791–800.

20 Bhatt DL, Bonaca MP, Bansilal S, et al. Reduction in Ischemic Events With Ticagrelor in Diabetic Patients With Prior Myocardial Infarction in PEGASUS-TIMI 54. J Am Coll Cardiol 2016; 67: 2732–40. 

21 Bonaca MP, Bhatt DL, Storey RF, et al. Ticagrelor for prevention of ischemic events after myocardial infarction in patients with peripheral artery disease. J Am Coll Cardiol 2016; 67: 2719–28.

22 Costa F, van Klaveren D, James S, et al. Derivation and validation of the predicting bleeding complications in patients undergoing stent implantation and subsequent dual antiplatelet therapy (PRECISE-DAPT) score: a pooled analysis of individual-patient datasets from clinical trials. Lancet 2017; 389: 1025–34.

23 Bhatt DL, Fox K, Harrington RA, et al. Rationale, design, and baseline characteristics of THEMIS: Effect of Ticagrelor on health outcomes in diabEtes Mellitus patients Intervention Study. Clin Cardiol 2019 Feb 20. doi: 10.1002/clc.23164.

24 Steg PG, Bhatt DL, Simon T, et al. Ticagrelor in patients with stable coronary disease and diabetes. Submitted to NEJM.

752753754

755756757758

759760

761762

763764765

766767768769

770771

772773774

775776777

778779780781

782783784785

786787

25 Bhatt DL, Cryer BL, Contant CF, et al. Clopidogrel with or without omeprazole in coronary artery disease. N Engl J Med 2010; 363: 1909–17.

26 Bhatt DL, Pollack CV, Weitz JI, et al. Antibody-based ticagrelor reversal agent in healthy volunteers. N Engl J Med 2019 Mar 17. doi: 10.1056/NEJMoa1901778. [Epub ahead of print]

27 Manoukian SV, Feit F, Mehran R et al. Impact of Major Bleeding on 30-Day Mortality and Clinical Outcomes in Patients With Acute Coronary Syndromes. An analysis from the ACUITY Trial. J Am Coll Cardiol 2007; 49: 1362–8.

28 Head SJ, Milojevic M, Daemen J, et al. Mortality after coronary artery bypass grafting versus percutaneous coronary intervention with stenting for coronary artery disease: a pooled analysis of individual patient data. Lancet 2018; 391: 939–48.

29 Bhatt DL. CABG the clear choice for patients with diabetes and multivessel disease. Lancet 2018; 391: 913–4.

30 Connolly SJ, Eikelboom JW, Bosch J et al. Rivaroxaban with or without aspirin in patients with stable coronary artery disease: an international, randomised, double-blind, placebo-controlled trial. Lancet 2018; 391: 205–18.

31 Anand SS, Bosch J, Eikelboom JW et al. Rivaroxaban with or without aspirin in patients with stable peripheral or carotid artery disease: an international, randomised, double-blind, placebo-controlled trial. Lancet 2018; 391: 219–29.

788789

790791792

793794795

796797798

799800

801802803

804805806

Figure 1: Kaplan-Meier event curves for the primary efficacy endpoint in patients with a history of PCI (top left panel), and with no history of PCI (top right panel)

807808

809

810

Figure 1 (continued): Kaplan-Meier event curves for TIMI major bleeding in patients with a history of PCI (bottom left panel), and with no history of PCI (bottom right panel)

811812813814

815816

817818

Figure 2: Kaplan-Meier event curves for net clinical benefit in patients with a history of PCI

819820

821

Figure 3: Hazard ratio for ticagrelor versus placebo as a function of time from the most recent PCI

Hazard ratios are calculated from a Cox proportional hazards model with treatment group, time since PCI, and interaction term for treatment group and time since PCI as explanatory variables. The solid line represents the estimated hazard ratio, and the dashes lines represent the 95% confidence intervals. The P-value of interaction between treatment group and time since PCI was 0.11.HR=Hazard ratio, PCI=Percutaneous Coronary Intervention.

822823824

825

826

Figure 4: Landmark analysis of the primary efficacy endpoint in patients with a history of PCI

827828829

830831

Figure 5: Number of events prevented or caused for every 1000 patients with a history of PCI treated for three years with ticagrelor

832833834

835